JPH0643848B2 - Rolling bearing abnormality diagnosis device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a rolling bearing abnormality diagnosing device used in various rotary machines, and more particularly to an AE (Acoustic).
Emission) measurement method is used.

[Conventional technology]

Many rolling bearings are used in the rotary sliding parts of rotating machines. If an abnormality (for example, seizure) occurs in the rolling bearings for any reason, it will cause the rotating machine to stop if it is significant. . Normally, a rotating machine is not installed alone, but is provided in an organically linked manner with other equipment, so that, for example, in a plant or the like, a situation in which the entire plant must be stopped may develop. Not rare. Therefore, it is necessary to detect abnormality of the rolling bearing early and accurately.

The following is known as a conventional rolling bearing abnormality detection technique. First, there is a known method using a so-called vibration measuring method in which vibration generated in a rotating machine is picked up (Japanese Patent Laid-Open No. 50-67184, Japanese Patent Laid-Open No. 52-123679, etc.). . They are,
In either case, the abnormality determination is performed by secondarily detecting a periodic sudden vibration waveform generated in the rotating machine due to the damage of the rolling bearing.

A second known example is one in which abnormality detection is performed using the AE measurement method (Japanese Patent Laid-Open No. 49-46982).

As a third known example, there is one that enables not only the determination of abnormality but also the location of damage occurrence in the rolling bearing (Japanese Patent Laid-Open No. 57-74627). However, this uses the vibration measurement method.

[Problems to be solved by the invention]

In the first known example, attention is paid to a sudden vibration waveform generated from a rolling bearing, and a bearing abnormality is determined by examining its level, periodicity, and the like. That is, when a scratch such as an indentation occurs on the ball, the roller, or the race surface, a sudden vibration waveform is generated, which is to be detected. The damage is advanced to some extent, and early detection of abnormality is difficult. On the other hand, when fine scratches such as wear and surface roughness of the balls, rollers, and race surfaces occur continuously, it is difficult to detect an abnormality by the above-mentioned conventional technique.

The second known example is evaluated in that the cause of the abnormality can be detected more directly by performing AE measurement, but it does not have an accurate diagnosis function such as the content of the abnormality or the identification of the abnormal portion.

In the third known example, the position can be specified with accuracy, and since it is still the vibration measurement method, it can be detected only after the vibration due to the abnormality of the rolling bearing occurs, and therefore the abnormality is detected early. I can't.

On the other hand, we conducted an experiment on the relationship between the cause of the abnormal phenomenon occurring in the rolling bearing by the AE measurement method and the signal form of the AE signal detected by the AE sensor. It was confirmed that either a continuous signal or a sudden signal was generated as the detection signal of the AE sensor. The present invention has been made based on such new findings.

An object of the present invention is to provide a rolling bearing abnormality diagnosing device capable of early detection of abnormality of each factor of a rolling bearing and determination of an abnormality occurrence location.

[Means for Solving Problems] An object of the present invention is to process an analysis result of waveform characteristics and frequency characteristics of a detection processing waveform of an AE signal generated due to an abnormality of a rolling bearing by a rolling bearing abnormality diagnosis algorithm. It is achieved by

That is, the present invention provides an AE sensor for detecting an AE signal generated from a rolling bearing, a detector for detecting the AE sensor output signal, and an average value level of the detector output signal and a normal level determined in advance. A signal level determiner for comparison, a waveform feature determiner for determining whether the detector output signal waveform is a continuous type or a sudden type, and a presence / absence of abnormality of the rolling bearing is determined based on the signal level determiner output signal. An abnormality determiner to do, the waveform feature determiner output signal shows a sudden type, and, when the abnormality determiner output signal indicates an abnormality, a frequency analyzer for performing frequency analysis of the detection output signal, Damage occurrence location for locating the damage occurrence location of the rolling bearing based on the rolling element passing frequency of the rolling bearing and the frequency of the maximum peak signal in the frequency analyzer output signal And position orientation device, is characterized in that and a display for displaying the abnormality determination result and damage occurrence location orientation results.

[Action]

The AE measurement method is used for early detection of abnormalities in rolling bearings. That is, when an abnormality occurs in the rolling bearing, the AE signal is generated even if the abnormality is minute.
For example, when minute defects such as surface roughness due to wear occur on the balls and rollers of the rolling bearing and the race surface, a continuous AE signal is generated. In addition, when a scratch such as an indentation occurs, even if the abnormality is in the initial state, a sudden AE
A signal is generated.

On the other hand, when the sudden AE signal is subjected to detection processing and the frequency component thereof is examined, the maximum peak frequency in the frequency component is found to be the frequency when the ball or roller passes through the race surface of the inner ring or outer ring, that is, the frequency of rolling element passage. Match.

In the present invention, early abnormality detection for each factor of the rolling bearing by examining the above waveform characteristics, and by comparing the maximum peak frequency with the rolling element passing frequency of the ball or roller, the location of the abnormality occurrence location can be determined. To do.

〔Example〕

 Next, an embodiment according to the present invention will be described with reference to the drawings.

FIG. 1 is a diagnostic algorithm for diagnosing a rolling bearing abnormality using the method according to the present invention, and FIG. 2 is an example of a system configuration for executing the diagnostic algorithm.

In FIG. 2, the AE sensor 3 is installed on the rolling bearing portion 2 installed on the rotary machine 1, the output of the AE sensor 3 is amplified by the amplifier 4, and the output is detected by the detector 5. Next, the output of the detector 5 is converted into a digital signal by an A / D converter, and then transferred to the signal analysis / evaluation unit 7. The signal analysis / evaluation unit 7 analyzes the waveform characteristics of the transferred detection waveform to diagnose abnormalities of rolling bearings according to factors, and displays the diagnostic results externally on a display unit 8 such as a CRT display.

Next, the details of the diagnostic processing performed by the signal analysis / evaluation unit 7 will be specifically described with reference to FIG. First, the detected signal 10 transferred to the signal analysis / evaluation unit 7 is subjected to the amplitude level determination by the waveform level determiner 11. Specifically, the level of the average value is calculated, and then the level is compared with the average value level when the rolling bearing is normally input to the signal analysis / evaluation unit 7 in advance. Is output.

On the other hand, the detected signal 10 is input to the waveform feature determiner 13,
It is determined whether the input detection signal is the continuous type or the sudden type. Looking at the AE signal generated due to the abnormality of the rolling bearing, as shown in FIG. 3, the type in which the amplitude level of the AE signal continuously rises with respect to the amplitude level of the signal at the normal time depending on the type of abnormality (Continuous type AE signal) and a type (abrupt type AE signal) in which the amplitude level of the AE signal suddenly rises. That is, by determining whether the AE signal is the continuous type or the sudden type, the factor of the rolling bearing abnormality can be determined.

Next, an example of a specific determination method of the waveform feature determination unit 13 will be described using the waveform shown in FIG.

In the waveform characteristic determiner 13, the average value voltage V of the digitally converted waveform after detection processing which is output from the A / D converter 6 is performed.
m is calculated from the arithmetic expression shown below.

Here, N is the sampling number set by the A / D converter 6, and V _i is the voltage value detected at each sampling number.

Next, the voltage value of the waveform exceeding V _m obtained by the above equation (1), that is, V _p1 , V _p2 , V shown in FIG.
_{From p3} ... _Vpi , the overall peak voltage average value _Vp is calculated from the following equation.

Next, using the above waveform analysis results, the continuous type and sudden type AE signals are determined by the following equation.

Here, K is a waveform determination constant, which is a sudden type when the formula (3) is satisfied and a continuous type when the formula (3) is not satisfied. The result determined by the above calculation is output to the abnormality determiner 12.

Next, the abnormality determiner 12 monitors the level ratio output from the level determiner 11, and when the level ratio exceeds the standard value, it determines that an abnormality has occurred in the rolling bearing. Next, to determine the cause of abnormality, the waveform feature determination function 1
Judgment is made from the continuous type and sudden type waveform feature determination results output from No. 3. That is, the average value level exceeds the standard value,
If the waveform is continuous, it is determined that damage is caused by continuous microscopic scratches due to wear or surface roughening.If the waveform is sudden, damage due to indentation or flaking is detected. It is determined that some have occurred. The above diagnostic results are transferred to the display unit 8 and the diagnostic results are displayed on the display unit 8.

According to this embodiment, in addition to early detection of abnormalities in rolling bearings,
Since the cause of the abnormality is also clarified, it is possible to prevent accidents such as feedback back to the optimum operating conditions.

Next, another embodiment of the present invention will be described with reference to FIGS. The present invention is an embodiment of an algorithm capable of locating a scratch occurrence point of a rolling bearing.

As shown in FIG. 5, when the abnormality determiner 12 damages the rolling bearing and the AE signal is diagnosed as a sudden type, the abnormality determiner 12 causes the FET processor 14 to operate.
An operation command is sent to. The FET processor 14 analyzes the frequency of the detected signal 10 based on the operation command. Next, the frequency analysis result is output to the flaw occurrence location locator 15 to diagnose the flaw occurrence location.

A specific method of locating the scratch occurrence location of the scratch occurrence location locator 15 will be described in detail with reference to FIGS. When the detection processing waveform of the sudden type AE signal generated due to the damage of the rolling bearing is frequency analyzed and the first peak frequency (F _p1 ) of the frequency is analyzed, as shown in FIG. Therefore, the first peak frequency value is different. The frequency due to the contact of rolling elements such as balls or rollers is calculated by the following formula. The rolling element passing frequency F _i at which one point on the inner ring contacts N balls or rollers is The rolling element passing frequency F _o at which one point on the outer ring contacts N balls or rollers is The rolling element passing frequency F _b at which one point on the ball or roller contacts the inner ring or the outer ring is Here, F _r is the rotation frequency of the bearing, d is the ball or roller diameter, and D is the pitch circle diameter.

FIG. 7 is based on the frequency analysis result shown in FIG.
It is an example of the calculation result which calculated | required each F _i , F _o , and F _b value from (4) (5) (6) Formula. As can be seen from FIG. 7, the rolling state in each element of the rolling bearing and the maximum peak frequency (F _p1 ) of the AE detection processing waveform described in FIG. 6 are in good agreement. That is, when each element rolls and passes through the wound, an AE signal is generated. In the flaw location locator 15, as shown in FIG. 5, the maximum peak frequency F _p1 and the calculation result F _i calculated by the rolling element passing frequency calculation means provided for calculating the above-described calculation formula are calculated. , F _o , F _b values are compared and the location of the scratch occurrence location is determined. Next, the result of orientation in the flaw location locator 15 is transferred to the display unit 8 and is transferred to the display unit 8.
The result is displayed at.

According to the present embodiment, since the abnormal portion of the rolling bearing becomes clear, it is possible to reduce the man-hours for maintenance.

〔The invention's effect〕

According to the present invention described above, since early detection of rolling bearing damage and determination of the damage factor and diagnosis of the location where the damage occurs are possible, accidents due to rolling bearing damage can be avoided, etc. Extremely large.

[Brief description of drawings]

FIG. 1 is a flow chart showing an algorithm of rolling bearing abnormality diagnosis according to an embodiment of the present invention, FIG. 2 is a system configuration diagram, and FIGS. 3 and 4 are explanatory diagrams for explaining the present invention. FIG. 5 is a flow chart showing a second embodiment of the present invention, and FIGS. 6 and 7 are explanatory views for explaining the second embodiment. 1 ... Rotary machine, 2 ... Rolling bearing, 3 ... AE sensor, 4 ... Amplifier, 5 ... Detector, 6 ... A / D converter, 7 ... Signal analysis and evaluation section, 8 ... Display Part, 11 ...
... Waveform level determiner, 12 ... Abnormality determiner, 13 ... Waveform feature determiner, 14 ... FFT processor, 15 ... Scratch location locator.

Front Page Continuation (72) Inventor Yuya Sato 4026 Kuji-cho, Hitachi City, Ibaraki Prefecture Hitachi Research Laboratory, Hitachi Co., Ltd. (56) References JP 61-189315 (JP, A) JP 55-54429 (JP, A) JP 56-135140 (JP, A) JP 60-31306 (JP, A)

Claims (1)

[Claims] 1. An AE sensor for detecting an AE signal generated from a rolling bearing, a detector for detecting the AE sensor output signal, and an average value level of the detector output signal and a normal level determined in advance. A signal level determiner, a waveform feature determiner for determining whether the detector output signal waveform is continuous or abrupt, and whether or not there is an abnormality in the rolling bearing based on the signal level determiner output signal An abnormality determiner, the waveform feature determiner output signal indicates a sudden type, and, when the abnormality determiner output signal indicates an abnormality, a frequency analyzer for performing frequency analysis of the detection output signal, the Damage occurrence location for locating the damage occurrence location of the rolling bearing based on the rolling element passing frequency of the rolling bearing and the frequency of the maximum peak signal in the frequency analyzer output signal Orientation device and the abnormality determination result and a display for displaying the damage occurrence point orientation results, the abnormality diagnosis device for the rolling bearing characterized by comprising a.